Research On Node Localization Algorithms In Anisotropic Wireless Sensor Networks

Wireless sensor network (WSN) is a kind of integrated intelligent informationprocessing platform, including sensor technology, micro-electromechanical technology,modern networking and wireless communication technology. Its development andapplication will have revolutionary consequences in every field of human life andproduction. Due to the sensor nodes’ uncontrollability when arranged, the majority of nodesin the network location can not be determined in advance while much applications of thewireless sensor network require nodes’ location information in the network to obtain theexact location of sources of information. So node localization is the basis of manyapplications of wireless sensor networks. Sensor nodes in the wireless sensor network aredivided into position-known nodes (anchor nodes) and position-unknown (unknown nodes).The unknown nodes take advantage of connectivity of the network, relationship between theneighbors and the distance to the anchor nodes to locate their positions. The main evaluationcriterions of position are the localization coverage and accuracy, the power consumptionand so on.This paper concludes and summarizes the existing localization systems and protocolsin WSN. Current localization approaches can be divided into two groups: range-based andrange-free. Due to the high costs and critical assumptions, the range-based schemes areoften impractical for WSNs. The existing range-free schemes, on the other hand, suffer frompoor accuracy and low scalability. One underlying assumption used in lot of theselocalization algorithms is that the network topology is isotropic. However, anisotropiccharacteristics which result from various factors such as the anisotropic terrain conditions,the different node densities and the irregular radio patterns, make the performances oflocalization algorithms get worse. Therefore, the research focus on the performance analysisof the localization algorithm in the anisotropy topology network which more close to theactual network, the principal tasks summarized as follows: 1、Doing analysis work on the structure of isotropic and anisotropic network topologyand the influence of the location algorithm. According to the characteristics of ranging-freelocalization algorithm, we analyze its performance in the two kind of networks, and carryout analysis and simulation about the existing algorithm by comparison.2、The localization algorithm based on the regional center affected by anisotropic is notserious from isotropic network, but the coverage of these algorithm is low, to this point, weput forward the solution to increase the localization coverage. To solve the problem aboutthe less neighbor anchor of some unknown nodes, the unknown nodes is treated as theanchor node after it was localized: if unknown nodes have neighbors anchor node, onlyusing neighbor anchor nodes achieve localized; on the other hand, if unknown nodes haveno neighbors anchor, use the neighbor unknown node which has been localized. Theimproved algorithm play a good role to improve the coverage but the accuracy of locationdecline at the same time, to this point, we put forward a scheme to increase the localizationaccuracy. the weighted centroid algorithm based on different communication radius isproposed, anchor node in this algorithm have two communication radius radio itsinformation packets, the unknown node give different weighted value to anchor neighborsand the assist anchor neighbors according to the two communication radius to improvelocalization accuracy.